"Mathematics is a powerful means to win arguments because people have a strong feeling that mathematics is objective, that "figures cannot lie." The truth is that it is easy to mislead and obfuscate a situation through the use of mathematical and statistical models that are inappropriate, whose assumptions are simplistic, or just plain wrong." -- mathematician W. ByersI'd planned to avoid this whole subject, but since Mike Collins has blanketed news sites with a press release of his Pearl River findings (in turn causing several emails to me), it seems increasingly unavoidable, especially since so many news venues are treating as new evidence, what is largely a rehash of previously-considered data. I encourage people, and especially scientists, to read Mike's article and view/listen to his video/audio tapes (and he has more material at his website), and decide for themselves....
The journal article is available for free here: http://tinyurl.com/3rutguj
You can go to this abstract page to link to some of Mike's referenced video/audio clips:
http://tinyurl.com/3h2uxx6
The current article is the latest culmination of a lot of reporting from Mike over the last few years beginning with his run-ins with folks at BirdForum.net. Let's just say that Mike has developed a propensity for publishing evidence with a degree of certainty that appears unwarranted (what William Byers would call "pseudo-objectivity through quantification").
His brief initial 2006 video did intrigue me a great deal at that time, and he reports several more sighting claims in the current article. However, lone self-report is never a strong basis for scientific conclusions, especially on a controversial subject. The strength of prior Cornell and Auburn reports was that they came from multiple observers in the same locale. Over the last dozen years many, many individuals have looked for IBWOs in the Pearl River area without success, while Mike keeps claiming encounters --- I grant that he has spent a tremendous amount of time there (and has access to certain areas that others don't have), and occasionally reports other individuals, under his tutelage seeing the species (although I've never seen a submitted full-account by any of those other individuals). And I admire Mike's dogged persistence, but given the public viewpoints he has espoused, Mike is now so vested in his own opinions/pronouncements, that his personal objectivity simply can't be assumed. Indeed, one of the flying Pearl River birds he filmed in 2009, turned out, on further analysis, to be a Red-headed Woodpecker (a bird not even half the size of an IBWO), despite his implication that it could be an Ivory-bill... perhaps indicative of the magnitude of a subjective bias.
But on to the current paper (and I'll only hit on a few points):
1) We now have 100s (perhaps 1000s) of reports of "kent" calls from deep woods, yet still no ability to definitively identify by recording (let alone by human ear) those kent calls that emanate from Ivory-billed Woodpeckers. All of Mike's verbal reports of kent calls are thus of little scientific value, although it is good when they come in conjunction with sightings, but those too are merely one-man verbal reports with no confirmation (and his recorded calls are, at best, inconclusive, no matter what sonograms show). The same fate befalls double-knocks, large cavities, and bark-scaling data --- these are interesting when found in close conjunction with good sightings, but by themselves remain weak evidence for Ivory-bills when presented in the literature. Many of Mike's "sightings" are cloaked in few details (how much time, what distance, what field marks, in what kind of light, etc.), again resulting in little more than shallow verbal claims (although he labels two sightings "of exceptional quality," whatever that means). His central focus in this paper seems to simply be on a few high-pitched calls and inconclusive double-knocks.
2) His discussion of the "high-pitched" calls, which he turns into putative Ivory-billed calls, is highly speculative, given that only a single brief reference by Tanner even hints that IBWOs ever make such sounds (and without a tape from Tanner there's no way of knowing if what Mike recorded even remotely resembles what Tanner heard). He then compares this unknown call to "putative kent calls" from Florida --- in other words he compares an unknown call with an unknown call and expects to derive some conclusion from that! (he does the same thing at a later point meaninglessly comparing a "putative double-knock" from the Pearl with a "putative double-knock" from Florida --- comparing two unknowns to each other is of little value; an unknown needs to be compared to a known/certain sample --- his related assumptions about the harmonic structure of IBWO calls, which he ties his sounds to, is based on a tiny sample size recorded on old equipment).
Also, the direction and distance of calls/sounds in deep woods can be notoriously difficult to establish or follow, but Mike assumes in one instance that he has followed the high-pitched calls accurately and can safely link them to a rapidly moving bird he presumes might, perhaps, possibly, in his judgment, be an Ivory-bill.
3) Mike's second video is his 2008 encounter with what appeared initially to be a Wood Duck, except that on further after-the-fact video analysis turns out, lo-and-behold, God-bless, to be an Ivory-billed Woodpecker! This is the bird that, according to Mike, revealed "a flap style that is radically different from what was expected for this species [IBWO]." Mike makes a big deal (as a new discovery) out of the fact that his bird folds its wings in-close during flight, claiming that old reports of the Ivory-bill having a "duck-like" flight style imply stiff-flapping (outstretched) wings... but this is simply erroneous. Previous historical accounts of "duck-like" flight for the IBWO are in reference to the fast, straight, direct, level flight path of the Ivory-bill (unlike the undulating flight of many woodpeckers), NOT its flap style, which could, like most woodpeckers, be presumed to include the closed wing periods --- absolutely NOTHING new or significant here --- in fact, Audubon reported the folded-in wings over 150 years ago, but Mike presents this as some grand NEW finding on his part!
4) In figure 4 Mike uses a a graph from Tobalske to compare data from a single Pileated Woodpecker to his 2008 video bird. For starters, using a single bird (with no context given) for comparison represents too small a sample size to be empirically meaningful, let alone conclusive, and I'd say that even if he was comparing his bird to a House Sparrow!! But even worse, the scales for the x-axis don't even match (time in ms) in the 2 graphs, making direct comparison difficult if not misleading (unless I'm missing something, and maybe I am). Further, no scales whatsoever are even presented for the y-axis, so no telling how different they may be!???.
Quantification, or application of mathematics, is a common practice to make weak data (or even junk science) appear scientifically stronger. Mike's videos are of poor quality and no amount of analysis changes that. If he is off, even a small amount, in any number of the initial measurements he calculates from these videos, than his end-point calculated numbers can be waaaay off. Additionally, Pileated flap rate and flight speed may vary from individual to individual and situation to situation (since we have little data), by whether it is a young bird or old, a gravid female, or a juvenile, or a cruising, fleeing, or chasing male, or an injured bird for that matter, or a bird dodging obstacles, or a startled bird, or flying into a head wind or pushed along by a tail wind... i.e. the variables are MANY, yet Mike's data set (and most data sets on both IBWO and PIWO) is very limited and simplistic. He no doubt thinks he can put a limiting range on such data for the PIWO, and then place his bird outside that range (even though we have no good flight data, beyond conjecture, for IBWO either); I don't think he can do that with available data.
One is left to wonder who, if anyone, peer-reviewed Mike's article, and if the editor of the acoustics journal it appears in has any significant comprehension of the subject matter.... (is it any surprise that no professional ornithological journal would touch this paper).
I'm skimming the surface here, but I think you get the point... the science presented is weak, and all the more-so, given no way to independently corroborate much of what Mike reports in print. Moreover, Mike's statements and behavior from the past have yielded him little credibility in much of the birding community (...and I understand that well because this blog has bestowed me with little credibility in much of the birding community!! ;-)), but he and I brought that upon ourselves; it is not the fault of others. As a result, no serious birding listserv, forum, or Website is much covering his recent publication (lest they be mocked for doing so), NOT even the Louisiana listserv!
As always, I believe it is possible that Ivory-bills may pass through the Pearl River region from nearby areas on occasion (I'm doubtful they breed or reside there), and I wish Mike luck in producing evidence that will demonstrate their presence to everyone's satisfaction. I'm not sure though that he comprehends what is necessary to accomplish that at this point; the bar is set very, VERY high (if he ever does document IBWOs conclusively he'll deserve immense gratitude from the birding community for incredible perseverance and hard work, but until then, well....). Over the decades 1000's of reports of Ivory-billed Woodpeckers have come in, the vast majority of which pretty quickly fall apart, and most of the rest never panning out upon follow-up. Time will tell, perhaps, if Mike's Pearl River claims are any different, or just more-more-more-of-the-same. For now, my Ivory-bill hopes reside elsewhere.
(p.s. -- if you're someone who hasn't seen the recent Pearl River IBWO news reports on the Web, and have no clue what this entire post is about... well, then... nnnnnnnnevermind!)
-----------------------------------------------------------------------
